You've spent months building your Obsidian vault. Daily notes, project pages, linked references, tags that actually mean something. It's your external brain for everything that matters at work.

Except meetings. The place where most decisions actually get made, where context gets shared, where someone says "let's do X" and three people nod — that all lives somewhere else entirely. A separate app, a separate silo, or more likely, nowhere at all.

You finish a call. Maybe you jot down two bullet points. Maybe you copy-paste a few lines from Otter into a daily note. But the full picture — who said what, the reasoning behind the decision, the action item someone volunteered for — evaporates within hours.

What if every meeting just appeared in your vault as a native markdown note? With YAML frontmatter you can query. Action items as checkboxes the Tasks plugin already sees. A full transcript you can search, link, and annotate. No export step, no copy-paste, no second app to check.

TL;DR

  • The gap: Meetings are your vault's biggest blind spot — decisions and context that exist nowhere in your PKM system
  • The fix: A recorder that writes native .md files directly into your vault folder
  • Bases & Dataview-ready: YAML frontmatter with title, date, duration, tags — queryable with built-in Bases or Dataview
  • Tasks-ready: Action items as standard - [ ] checkboxes, bidirectional sync
  • Semantic search: Ask questions across all meetings by meaning, not just keywords

The Gap in Your Vault

If you're serious about Obsidian, you've probably solved most of your knowledge capture problems. Articles go through a read-it-later pipeline. Highlights sync automatically. Project notes link to reference material. Daily notes tie everything to a timeline.

But meetings are the one source that resists this workflow. And it's not because they're unimportant — it's because the tools that capture them don't speak your language.

Cloud meeting recorders like Otter, Fireflies, and Fathom store your transcripts behind a login, in their own format, in their own silo. You can export, sometimes, but it's always a second-class experience. A PDF. A plain text dump. Something you'd have to reformat before it's useful in your vault.

Meanwhile, the information trapped in those tools is often the most valuable kind you produce at work. Not the article you bookmarked — the actual conversation where your team debated the trade-offs, made the decision, and assigned the follow-ups. That's original, first-hand knowledge. And you're letting it sit in a silo your vault can't reach.

What a Meeting Note Looks Like in Your Vault

Imagine you finish a Zoom call. A few seconds later, this file appears in your vault:

--- title: Weekly Standup date: 2026-04-04T10:00:00 duration: "23:15" tags: - meeting - project-alpha language: en status: summarized --- ![](_audio/2026-04-04%20Weekly%20Standup.mp3) ## Summary The team reviewed progress on the API migration. Sarah confirmed the database schema changes are complete. Marcus raised a concern about backwards compatibility with v2 clients. ## Action Items - [ ] Draft v2 deprecation timeline - [ ] Update API docs for v3 endpoints - [x] Schedule stakeholder review ## Transcript **You** (0:00:00) Let's start with the roadmap updates. The schema migration landed yesterday. **Other** (0:00:43) Great. I have a concern about backwards compat though. **You** (0:01:12) Good point. Let's set a deprecation timeline.

That's a real Obsidian note. You can tag it, link it to your project page, add your own annotations between sections. The audio file embeds as a playable audio widget — click play right inside the note to hear the exact tone of a discussion.

The file structure in your vault is clean:

My Vault/ Meetings/ ← point mono at this folder 2026-04-04 Weekly Standup.md 2026-04-03 Client Call.md 2026-04-01 1-on-1 with Sarah.md _audio/ 2026-04-04 Weekly Standup.mp3 2026-04-03 Client Call.mp3 2026-04-01 1-on-1 with Sarah.mp3

Flat files, sorted by date naturally. Audio tucked away in _audio/ so your sidebar stays clean. No database, no proprietary format. Just markdown and mp3.

Queryable with Bases and Dataview

Every meeting note has standard YAML frontmatter: title, date, duration, tags, status. The same properties you'd put in any well-structured vault note. That means two powerful ways to query them.

Obsidian Bases (built-in, no plugins)

Obsidian's built-in Bases feature turns your meetings folder into a database-like view — no code, no community plugins, nothing to install.

Create a new Base, point it at your Meetings folder, and you get a sortable, filterable table of every meeting with columns for date, duration, tags, and status. Click a cell to edit a property inline — add a tag, change a status — and it writes directly to the note's frontmatter.

Filter to just this week's meetings. Sort by duration. Group by tag. All through a visual interface that feels like a spreadsheet but operates over your actual markdown files.

If you've used Notion databases, this is that — but over files you own.

Dataview (for power users)

If you prefer queries in code, Dataview works just as well:

All meetings this week tagged #project-alpha:

```dataview TABLE duration, status FROM "Meetings" WHERE contains(tags, "project-alpha") AND date >= date(today) - dur(7 days) SORT date DESC ```

Meetings longer than 30 minutes you haven't summarized yet:

```dataview LIST FROM "Meetings" WHERE status = "transcribed" AND duration > "30:00" SORT date DESC ```

Both approaches work because the underlying data is the same: standard YAML frontmatter. Use Bases for quick visual browsing, Dataview for complex queries embedded in your project pages. Or both.

Action Items That Work with Obsidian Tasks

This is the part that makes the biggest difference in practice.

Meeting action items are standard markdown checkboxes under a ## Action Items heading. The format the Tasks plugin already queries natively:

- [ ] Draft v2 deprecation timeline - [ ] Update API docs for v3 endpoints - [x] Schedule stakeholder review

One Tasks query to see every uncompleted action item from every meeting this month:

```tasks path includes Meetings not done ```

Check off an item in Obsidian. The recorder app sees the change. Check it off in the recorder app. Obsidian sees the change. Bidirectional — because it's the same file.

No more action items buried inside AI-generated summaries you never re-read. No more "wait, who was supposed to do that?" No more searching through three different tools to find what you committed to last Tuesday.

Why checkboxes, not YAML arrays: Obsidian Tasks queries - [ ] lines. YAML frontmatter arrays are invisible to it. By keeping action items in the body as checkboxes, they work with Tasks out of the box — no configuration, no workarounds.

Embedded Audio Playback

Each meeting note embeds its audio file with Obsidian's standard image/file embed syntax:

![](_audio/2026-04-04%20Weekly%20Standup.mp3)

Obsidian renders this as a playable audio widget right inside the note. You're reading the transcript, you hit a section where tone matters — a disagreement, a hesitation, a joke that changed the direction of the conversation — and you click play right there.

Audio files live in a _audio/ subfolder. The underscore prefix sorts it to the top and out of the way. Your sidebar shows clean, date-sorted note titles, not a mix of markdown and mp3 files.

Semantic Search Across All Your Meetings

Obsidian's built-in search is good for finding exact words. But meetings are conversational — the same concept gets described in different ways across different calls.

You remember a discussion about pricing strategy, but the transcript says "revenue model." You know someone raised a concern about the timeline, but they phrased it as "I'm not sure we can ship by Q3." Keyword search misses these. Semantic search doesn't.

Behind the scenes, the recorder chunks your transcripts and embeds them into a vector database. When you ask a question, it retrieves the relevant chunks by meaning, not by exact string match:

This complements Obsidian's search rather than replacing it. Obsidian finds the notes; semantic search finds the moments within them.

And because the vector index is stored in a hidden .mono/ folder inside your meetings directory, it's invisible to Obsidian. Delete it, and it gets rebuilt from your markdown files. Your notes are always the source of truth — the search index is just a cache.

Your Files, Not Ours

This is the part that matters most if you've been burned by a tool shutting down, changing pricing, or holding your data hostage.

Every meeting is a plain .md file and an .mp3 file. That's it. No database you can't open. No proprietary format you can't read. No cloud account that gates access to your own recordings.

Compare that to what you get from cloud meeting recorders:

Cloud recorders Markdown-first
Your data lives... On their servers On your disk
Export format PDF, plain text, or nothing Already markdown
Obsidian integration Manual copy-paste Native — files appear in your vault
Bases / Dataview / Tasks Not possible Works out of the box
If you stop paying Scramble to export before lockout Nothing changes — files are yours
If the app disappears Your recordings may too You still have .md + .mp3 files

Uninstall the recorder app tomorrow and you lose nothing. Every summary, every action item, every transcript is right there in your vault as a file you can open in any text editor on any operating system.

Edits Go Both Ways

The recorder watches your meetings folder for changes. This means edits you make in Obsidian are picked up automatically:

You're never locked into editing in one place. Use whichever tool is in front of you.

Your own notes are safe too. Add paragraphs between sections, drop links to project pages, insert block references. The recorder preserves any content it didn't create — your annotations survive when the file gets updated.

How to Set It Up

Three steps:

  1. Create a folder in your vault for meetings — Meetings/, Calls/, whatever fits your structure
  2. Point mono's recordings folder to that folder in settings
  3. Record a meeting — a .md file appears in your vault when transcription finishes

From there, the workflow is automatic. Every new recording produces a note with frontmatter, AI-generated summary, action items, and full transcript. Tag it, link it to your project, and move on.

mono handles the recording, transcription, and summarization. It runs on Windows and Mac, records audio from any app (Zoom, Teams, Meet, Discord, anything), and does transcription locally on your machine with Whisper. No audio leaves your device. No cloud account required.

Already have recordings? Existing recordings from before the Obsidian integration will be migrated automatically in a future update. For now, new recordings go straight into your vault.

What Your Vault Looks Like After a Month

After a few weeks, the compound effect kicks in. Your vault goes from "notes plus a separate meeting tool" to a single system where everything connects.

Your daily note links to today's meetings. Not manually — because the meeting note has today's date in frontmatter, a simple Dataview query in your daily template shows them automatically.

Your project page surfaces every relevant meeting. Tag meetings with a project tag, and a Dataview table on the project page lists them chronologically with duration and status.

Your weekly review includes action items from every call. One Tasks query aggregates them all. Nothing slips through the cracks because nothing lives outside the vault anymore.

Onboarding a new team member? Share the vault folder. They get every meeting transcript, every decision with its reasoning, every action item with its status. Months of context in seconds.

The meeting silo is gone. Conversations, notes, projects, and action items all live in one place, linked by the connections you've already built.

Your meetings belong in your vault

mono records any meeting app, transcribes locally with Whisper, and writes native markdown files directly into your Obsidian vault. One recording free, no account needed.